HLA class II molecules are peptide-binding glycoprotein dimers. Polymorphic residues within class II molecules influence the avidity, specificity, and conformation of peptide binding interactions. These same polymorphisms discriminate HLA class H genes which are associated with IDDM from those which are not. It is highly likely that specific peptide- class II recognition is critical for pathogenic events in this disease, and that the class H molecules which are involved are those which are genetically associated with disease susceptibility, namely, specific HLA- DQ dimers. In this project we will identify and characterize the detailed topography of the peptide-binding site in such class H molecules, and modulate immune recognition of the peptide-class II complex by alterations of specific peptide residues. Specific Aims are: Aim 1. To identify key structural determinants of peptide orientation and anchoring within the HLA-DQ peptide-binding groove. Aim 2. To construct a binding site profile for the DQ3.2 molecule associated with IDDM. Aim 3. To characterize peptide analogs which interfere with or modulate specific HLA-DQ-dependent peptide recognition. A combined molecular modeling and peptide binding strategy will be used, in concert with functional assays of T cell recognition. Derivation of a detailed binding site profile for HLA molecules and candidate autoantigens will aid our understanding of early pathogenic events in IDDM and of structural requirements for a new category of novel immunotherapeutics